To allow for an operationalization of agenda bias, I use parties’ campaign communication as an approximation of the potential universe of news stories (D’Alessio & Allen, 2000; Eberl, 2017). I compare the policy issues addressed in campaign communication (i.e., the party agenda) with the policy issues the parties address in media coverage (i.e., the mediated party agenda).

To discover the latent topics in the corpus of press releases (1.942) and news articles (11.880), a structural topic modeling (STM) developed by Roberts (2016) is applied. The STM is an unsupervised machine learning approach that models topics as multinomial distributions of words and documents as multinomial distributions of topics, allowing to incorporate external variables that effect both, topical content and topical prevalence.

Structural Topic Model

Build Corpus

Select Model

STM assumes a fixed user-specified number of topics. There is not a “right” answer to the number of topics that are appropriate for a given corpus (Grimmer and Stewart 2013), but the function searchK uses a data-driven approach to selecting the number of topics. The function will perform several automated tests to help choose the number of topics including calculating the held out likelihood (Wallach et al. 2009) and performing a residual analysis (Taddy 2012).

## stm v1.3.3 (2018-1-26) successfully loaded. See ?stm for help. 
##  Papers, resources, and other materials at structuraltopicmodel.com

Run Model

I included the document source as a control for the topical topical prevalence and the type (press release or news article) as a control for topical content. Thus, I assume that the distribution of topics depends on the source and the word distribution within each topic differes between party press releases and news articles. The number of topics is set to 55.

Results

Label topics

To explore the words associated with each topic we use the words with the highest probability in each topic. As we included the source type (press release or news paper) as a control for the topical content (the word distribution of each topic), we have two different labels for each topic.

topic_name_news topic_name_press
1 Topic 1 : eu europa deutschland europäischen Topic 1 : eu europäischen europäisch kommiss
2 Topic 2 : spd schulz gabriel nahl Topic 2 : gabriel außenminist amt sigmar
3 Topic 3 : koalit spd jamaika neuwahlen Topic 3 : koalit große großen jamaika
4 Topic 4 : wahl bundestagswahl parteien mitglied Topic 4 : gewählt europarat parlamentarischen deutschen
5 Topic 5 : vw kommunen diesel konzern Topic 5 : deutschland auto theurer energien
6 Topic 6 : auto diesel hendrick deutschland Topic 6 : diesel fahrverbot bundesregierung autoindustri
7 Topic 7 : the of to on Topic 7 : school world max up
8 Topic 8 : trump gipfel hamburg donald Topic 8 : usa trump un präsident
9 Topic 9 : land steinmeier deutschland bundespräsid Topic 9 : zukunft bildung deutschland stärken
10 Topic 10 : grünen csu jamaika cdu Topic 10 : verbrauch verbraucherschutz verbraucherinnen ernährung
11 Topic 11 : afd petri partei frauk Topic 11 : sachsen anhalt poggenburg andré
12 Topic 12 : prozent spd umfrag afd Topic 12 : prozent stimmen arbeitsgrupp vergleich
13 Topic 13 : link linken rot partei Topic 13 : asylpolitik nahl andrea film
14 Topic 14 : spd union groko schulz Topic 14 : bundestagsfrakt sprecher hierzu koalitionsvertrag
15 Topic 15 : niedersachsen cdu landtag grünen Topic 15 : de bundesvorstandsmitgli paul maizièr
16 Topic 16 : maizièr thoma verfassungsschutz daten Topic 16 : daten grundrecht bürger stellvertretend
17 Topic 17 : euro milliarden union geld Topic 17 : euro milliarden jahr einkommen
18 Topic 18 : kohl helmut kanzler jahr Topic 18 : ost osten deutschen einheit
19 Topic 19 : auftritt rede menschen bühne Topic 19 : demokrati besuch whistleblow menschen
20 Topic 20 : frauen männer urteil männern Topic 20 : frauen gewalt türkei hilf
21 Topic 21 : fdp lindner jamaika grünen Topic 21 : demokraten lindner freien dr
22 Topic 22 : csu seehof horst union Topic 22 : forderungen forderung alexand wähler
23 Topic 23 : afd partei wähler parteien Topic 23 : partei facebook demokratisch urteil
24 Topic 24 : august spd cdu prozent Topic 24 : deutschland antwort wahl bundestag
25 Topic 25 : stiftung deutschland osten menschen Topic 25 : deutschland menschen forschung wirtschaft
26 Topic 26 : zdf illner sendung sender Topic 26 : öffentlich rechtlichen rundfunk medien
27 Topic 27 : spahn jen cdu heimat Topic 27 : frage berger gesellschaft antisemitismu
28 Topic 28 : merkel angela kanzlerin cdu Topic 28 : merkel angela bundeskanzlerin kanzlerin
29 Topic 29 : schmidt eu glyphosat deutschland Topic 29 : eu türkei glyphosat bundesregierung
30 Topic 30 : afd höcke partei björn Topic 30 : meuthen jörg bundesvorstand österreich
31 Topic 31 : maa gesetz heiko justizminist Topic 31 : de schäubl masi banken
32 Topic 32 : flüchtling familiennachzug flüchtlingen deutschland Topic 32 : menschen familiennachzug deutschland engag
33 Topic 33 : afd weidel gauland alic Topic 33 : deutschland weidel alic migranten
34 Topic 34 : welt menschen sonntag frage Topic 34 : bund schulen kommunen länder
35 Topic 35 : kinder bildung schulen bund Topic 35 : kinder arbeit fraktion mindestlohn
36 Topic 36 : bundestag afd abgeordneten abgeordnet Topic 36 : bundestag antrag deutschen fraktion
37 Topic 37 : spd partei koalit große Topic 37 : agenda sozial digitalisierung digital
38 Topic 38 : fragt bosbach leut talk Topic 38 : recht rechtsstaat rechten extremismu
39 Topic 39 : hamburg gipfel polizei scholz Topic 39 : hamburg polizei gipfel linksextremismu
40 Topic 40 : grünen özdemir göring eckardt Topic 40 : grünen erklären katrin hofreit
41 Topic 41 : berlin amri anschlag polizei Topic 41 : berlin beschäftigten air bundesregierung
42 Topic 42 : abschiebungen abschiebung deutschland afghanistan Topic 42 : pflege gesetz menschen bundesregierung
43 Topic 43 : schulz spd martin duell Topic 43 : schulz martin deutschland kanzlerkandidat
44 Topic 44 : bundeswehr leyen soldaten nato Topic 44 : bundeswehr bundesregierung nato syrien
45 Topic 45 : deutschland antisemitismu polizei israel Topic 45 : menschen deutschland opfer tag
46 Topic 46 : türkei erdogan gabriel deutschland Topic 46 : russland sanktionen russischen außenpolitisch
47 Topic 47 : staatsanwaltschaft ermittlungen verdacht vorwürf Topic 47 : jelpk ulla innenminist fraktion
48 Topic 48 : wagenknecht bartsch kip sahra Topic 48 : fraktion hänsel bundesregierung vorsitzend
49 Topic 49 : jahr zahl deutschland bamf Topic 49 : bundesregierung deutschland jahr erdogan
50 Topic 50 : post abstimmung entscheidung rheinischen Topic 50 : gesetz gesetzentwurf bundesrat maut
51 Topic 51 : hau brandt willi spd Topic 51 : kerstin johann michel brandt
52 Topic 52 : ge ten be ver Topic 52 : deutsch macron zusammenarbeit frankreich
53 Topic 53 : cdu schleswig altmaier günther Topic 53 : religionsfreiheit menschenrecht bundestagsfrakt weltweit
54 Topic 54 : söder csu bayern herrmann Topic 54 : bayern bayerischen bayerisch harald
55 Topic 55 : politik wolfgang jahren politisch Topic 55 : politik politisch jahren frage

Topic frequency

The expected proportion of the corpus that belongs to each topic is used to geht an initial overview of the results. The figure below displays the topics ordered by their expected frequency across the corpus. The four most frequent words in each topic are used as a label for that topic.

Measure Agendas

For each document, we have a distribution over all topics:

What is the document acutally about?

source title_text
1 focus.de

Die Linke im Streit: Bizarrer Machtkampf in der Führungsriege Wagenknecht, Bartsch, Kipping und Riexinger - in der Führungsriege der Linken geht es nicht gerade harmonisch zu. Der Burgfrieden des Wahlkampfs ist dahin. Sorgen um die Stammwähler sind groß.

Eigentlich kann die Linke froh sein, dass sie als kleinste Oppositionskraft hinter SPD und AfD im Moment im Windschatten der beginnenden Jamaika-Sondierungen segelt. Trotz schmerzhafter Wählerverluste im Osten leistet sie sich aber einen bizarren Machtkampf.

Am Dienstag und Mittwoch kommt die neu gewählte Linken-Fraktion in Potsdam zur Klausur zusammen. Sahra Wagenknecht und Dietmar Bartsch wollen wieder an die Spitze gewählt werden. Doch unmittelbar vor dem Treffen gibt es noch Gerangel. So liegen Anträge vor, nach denen die in der Fraktion vertretenen Parteichefs Katja Kipping und Bernd Riexinger im Fraktionsvorstand mitbeschließen dürfen sollen. Gestritten wird auch darüber, wer im Plenum des Bundestags wie für die Linke sprechen darf.

Es rächt sich bei der Linken, dass Konflikte unter den Teppich gekehrt wurden. So rief die Partei im Dezember Wagenknecht und Bartsch erst zu den Spitzenkandidaten im Wahlkampf aus, nachdem beide klar gemacht hatten, dass sie nicht für ein Spitzenquartett mit den Parteichefs Kipping und Riexinger zu haben sind. Nach der Wahl scheint der Burgfriede dahin.

Zunächst sorgte Bundesgeschäftsführer Matthias Höhn für Schlagzeilen. Er wolle wegen eines Zerwürfnisses mit Kipping und Riexinger zurücktreten, hieß es. Er habe die beiden nicht öffentlich gegen die Kritik von Oskar Lafontaine in Schutz genommen, sollen Kipping und Riexinger Höhn vorgeworfen haben. Lafontaine, der Mann Wagenknechts, hatte bei Facebook unter anderem gepostet, dass die Parteichefs “wenig Zustimmung bei den Wählern” haben.

Agendas were measured in terms of percentage distributions across the 40 topics. For each source the average distribution of each topic is calculated.

Then, we estimated bivariate correlations between party agendas and the mediated party agendas in the online news. These correlations represent the agenda selectivity each party experiences in each media outlet. The higher the correlation, the more congruent both agendas are (Brandenburg, 2005; Harris, Fury, & Lock, 2006; Kim, Xiang, & Kiousis, 2011).

Again, to measure the bias and not just outlet specificities, for each outlet the mean agenda selectivity of all other parties was subtracted from each party’s specific agenda selectivity value and then standardized to range from −1 to 1, where −1 stands for both agendas being not congruent at all and +1 stands for both agendas being identical.

This underlines the importance of studying all three types of media biases simultaneously, as examining just one aspect provides a misleading picture of the extent and nature of bias. The existence of various and diverse forms of media bias also means that it is worth considering their distinct effects on party preferences.